It was a snowy late-November morning at Chamberlain, South Dakota (9V9). A Pilatus PC-12 had sat out on the ramp during a night of intermittent snowfall and freezing drizzle. Its passengers, who had flown in from Idaho Falls the previous day to hunt pheasants, planned to return home that day.
While the rest of the party was out shooting, the private pilot, 48, and one companion got some isopropyl alcohol de-icing fluid from a hardware store, borrowed a ladder from the hunting lodge at which they had stayed, and spent three hours chipping snow and ice from the wings. The ladder was not tall enough to allow them to reach the upper surface of the T-tail, but the pilot was satisfied that the rest of the airplane was sufficiently clean.
Video of the Pilatus taxiing out showed snow falling heavily and white clumps adhering to parts of the fuselage and vertical tail. A couple of inches of snow (and presumably some ice) lay on the top of the horizontal stabilizer. The takeoff was recorded as well. The Pilatus roared down Runway 31, lifted off, banked to the left, and faded from sight in the snow and mist.
No one at the airport knew it at the time, but it crashed less than a mile from the runway. Of the 12 people aboard, three survived with serious injuries. The pilot was among the nine dead.
Thirty years ago, it would have looked like an open and shut case. Whatever residue of ice remained on the wings must obviously have triggered a premature stall. But we live in a different era now, with flight data and cockpit voice recorders in wide use. They tell accident investigators not what must have happened, but what really did.
The National Transportation Safety Board’s probable cause finding made no mention of snow and ice. It attributed the loss of control after takeoff and the ensuing stall to “the pilot’s improper loading of the airplane, which resulted in reduced static longitudinal stability.” Another contributing factor was “his decision to depart into low instrument meteorological conditions”—although that seems unfair, since the whole point of having an instrument rating and a powerful airplane equipped for flight in known icing is to be able to do exactly that.
The cockpit voice recorder picked up the sounds of passengers boarding the airplane, stomping snow from their shoes, clicking their seat belts. One passenger commented on how many pheasants they had bagged. Another recited a prayer of gratitude for various blessings—it was Thanksgiving weekend—and went on, with eerie prescience, “Father in Heaven, we ask for a special blessing now that we take off in this not-so-great weather and that [Thou wilt] watch over and protect us. Impress upon the mind of [the pilot] that he might know how best to travel this course that we are about to do, and we are thankful for this airplane and ask that You will watch over and protect us.” A collective “Amen” followed.
The pilot and the right-seat occupant radioed the airport manager, who was plowing the runway, to ascertain its condition. Their exchange was somewhat acerbic. The manager frankly told the pilot he must be crazy. The pilot good-naturedly replied that the snow berms on either side of the plowed portion of the strip were not a concern. As it turned out, he was right.
The pilot back-taxied to the approach end of Runway 31 and succeeded in turning the airplane around. The power came up, the Pilatus accelerated, and after 30 seconds it rotated. The pitch angle increased to almost 20 degrees, then eased back to about 10. Practically from the moment of liftoff, the stall warning sounded and an automated voice intoned the word “stall” over and over, no fewer than 19 times. Eleven seconds after rotation, a porpoising motion began, increasing in magnitude and rapidity. The bank angle increased to 64 degrees; the stick pusher actuated and, at a height of 380 feet, the Pilatus stalled.
With granular information from the flight data recorder, the NTSB conducted simulations to ascertain whether the airplane had been controllable and whether the accumulations of snow and ice remaining on it could have been a factor in the accident. The conclusion was that the airplane should have been controllable, and that the snow and ice had not significantly degraded its performance, though they may have affected the elevator control forces.
The data recorder stored a number of previous flights, and the NTSB noted that the pilot, who had 1,260 hours in type, habitually rotated somewhat abruptly, tending to slightly overshoot the desired pitch attitude and then correct. Another pilot who regularly flew the airplane used a gentler, more gradual rotation, which the board found made speed control easier.
The board compared the accident flight with the previous day’s trip from Idaho Falls to Chamberlain. The cabin loading had been similar, and there were pitch oscillations after takeoff on that flight as well. The crux of the matter, in the NTSB’s view, was the combination of heavy weight—the airplane was 107 pounds over gross—and the CG location, several inches behind the aft limit, that resulted from 12 people, none of them lap children, and a great many dead pheasants occupying a 10-passenger airplane. An aft CG is associated with diminished stick forces and weak speed stability, conditions that may be difficult to manage on instruments.
The stall warnings that were heard practically from the moment the airplane rotated were due to the design of the Pilatus’s ice protection system. When ice protection is on, the triggering speeds for both the stall warning and the stick pusher increase considerably. According to the flight manual, the target rotation speed at max gross in icing conditions was 92 knots. The pilot rotated at 88, possibly because he wanted to get clear of snow build-up on the partially plowed runway. When the actual stall occurred, however, the indicated airspeed was only 80 knots.
One can speculate about what passed through the pilot’s mind during the few seconds between the liftoff and the stall. The aural stall warning must have taken him by surprise. Since he had just spent hours removing snow and ice, his first thought may have been that it was caused by some lingering contamination on the wings. But now he was in near-whiteout conditions, and too low to risk pushing the nose down decisively. The airplane may not have responded to a gentle push on the yoke. Pitch oscillations made speed control difficult. There was little time to analyze or adapt—only enough for an exclaimed “Oh no!”
The pilot was the kind of person whom you would expect to follow rules. Yet he ignored the CG limits. Did he feel undue pressure to get his passengers back home? Probably not. There is no indication that he hesitated or considered the takeoff dangerous; in fact, he seemed less concerned than his prayerful passengers were. Did he understand how the extreme aft loading could affect the airplane’s flying qualities? He had made a similar flight the day before. Did he begin this one thinking it would be exactly the same?
Sometimes you don’t know how near the edge you are until you go over it.
This article is based on the National Transportation Safety Board’s report of the accident and is intended to bring the issues raised to our readers’ attention. It is not intended to judge or to reach any definitive conclusions about the ability or capacity of any person, living or dead, or any aircraft or accessory.
